VIRUSES > 397
Filtration. As the original designation of these agents indicates,
filtration has been extensively employed in their study. It. was
early recognized that an important criterion of the viruses was
their ready passage through bacterial filters. This criterion still
holds for general application, but certain reservations are neces-
sary. The process of, and the factors involved in, filtration have
already been discussed (p. 67). It is sufficient now to point out
that filtration is not a simple mechanical action governed only
by the relative sizes of the pores of the filter and the particles to
be filtered. Many factors are concerned and the most important
are : the composition and electrical charge of the filter ; the
plEL and electrical charge of the material undergoing filtration ;
the amount of extraneous material in the suspension; the
duration of filtration ; the pressure employed ; the temper at me
at which it is carried out. Viruses may be adsorbed on to the
filter or on to albuminous material in the suspension and so fail to
pass through a filter, which ought to be traversed. Irregular
results have not infrequently been due to some variation in the
filtration technique. The value of filtration in the study of
viruses is therefore limited. The process cannot be used to
determine with any accuracy the size of the virus particles, but
by the use of the various grades of filter a means of obtaining a
rough estimate of their dimensions is available.
Many viruses readily pass through the usual bacterial filters,
such as Berkefeld " N " and Seitz " EK " types ; others, e.g.,
the vaccinia and herpes viruses, only traverse with difficulty the
coarser types, such as the Berkefeld " V ", when special pre-
cautions are taken. Some smaller bacteria and various spiro-
chsetes may pass through the coarser filters more readily than the
larger viruses. It consequently follows that filtration alone is
not an absolute criterion for the identification of a virus.
Ultra-filtration. More precise studies of the size of different
viruses have recently been made by the use of ultra-filtration,
which was first employed for the study of colloidal suspensions
by Bechhold (1907). This process involves the use of collodion
membranes, which are less influenced by outside factors than the
bacterial filters. Much of the early application of these mem-
branes to the study of the viruses was unreliable owing to the
use of faulty membranes, but in recent years a successful technique
has been introduced by Elford. Some of the main points of this
are : the grading of acetic acid membranes in terms of their
average pore diameters, constant area- of filtering surface, use of